Pain measurement in children is poorly documented. Probably the most reliable method uses a visual analog scale involving the marking of a point in a line representing extremes from "no pain" to "severe pain, the most possible". Again, unpleasantness can be described in words, but is perhaps best quantified using a visual analog scale with extremes denoting "no unpleasantness" to "extremely unpleasant" for adults, or the selection from a series of picture faces ranging from smile to crying in the case of children (after Rogers, et al.). Although nitrous oxide has been used anecdotally for analgesia during diagnostic procedures in children with cancer, no detailed series of studies are published in the literature. We propose to perform a pilot study looking at the feasibility of administration and efficacy of nitrous oxide in these patients.